WCK 5107 (Zidebactam) and WCK 5153 Are Novel Inhibitors of PBP2 Showing Potent “β-Lactam Enhancer” Activity against Pseudomonas aeruginosa, Including Multidrug-Resistant Metallo-β-Lactamase-Producing High-Risk Clones

ABSTRACT

Zidebactam and WCK 5153 are novel β-lactam enhancers that are bicyclo-acyl hydrazides (BCH), derivatives of the diazabicyclooctane (DBO) scaffold, targeted for the treatment of serious infections caused by highly drug-resistant Gram-negative pathogens. In this study, we determined the penicillin-binding protein (PBP) inhibition profiles and the antimicrobial activities of zidebactam and WCK 5153 against Pseudomonas aeruginosa, including multidrug-resistant (MDR) metallo-β-lactamase (MBL)-producing high-risk clones. MIC determinations and time-kill assays were conducted for zidebactam, WCK 5153, and antipseudomonal β-lactams using wild-type PAO1, MexAB-OprM-hyperproducing (mexR), porin-deficient (oprD), and AmpC-hyperproducing (dacB) derivatives of PAO1, and MBL-expressing clinical strains ST175 (bla_VIM-2) and ST111 (bla_VIM-1). Furthermore, steady-state kinetics was used to assess the inhibitory potential of these compounds against the purified VIM-2 MBL. Zidebactam and WCK 5153 showed specific PBP2 inhibition and did not inhibit VIM-2 (apparent K_i [K_{i app}] > 100 μM). MICs for zidebactam and WCK 5153 ranged from 2 to 32 μg/ml (amdinocillin MICs > 32 μg/ml). Time-kill assays revealed bactericidal activity of zidebactam and WCK 5153. LIVE-DEAD staining further supported the bactericidal activity of both compounds, showing spheroplast formation. Fixed concentrations (4 or 8 μg/ml) of zidebactam and WCK 5153 restored susceptibility to all of the tested β-lactams for each of the P. aeruginosa mutant strains. Likewise, antipseudomonal β-lactams (CLSI breakpoints), in combination with 4 or 8 μg/ml of zidebactam or WCK 5153, resulted in enhanced killing. Certain combinations determined full bacterial eradication, even with MDR MBL-producing high-risk clones. β-Lactam–WCK enhancer combinations represent a promising β-lactam “enhancer-based” approach to treat MDR P. aeruginosa infections, bypassing the need for MBL inhibition.

INTRODUCTION

Pseudomonas aeruginosa is a major opportunistic human pathogen that frequently causes severe nosocomial infections (1–3). In the clinic, P. aeruginosa is characterized by the remarkable ability of acquiring and expressing multiple resistance mechanisms, thereby becoming recognized as one of the most difficult-to-treat multidrug-resistant (MDR) pathogens (3–5). The acquisition of potent exogenous β-lactamases, such as metallo-β-lactamases (MBLs) or extended-spectrum β-lactamases (ESBLs), through horizontal gene transfer is an endemic growing threat (6–10). However, the mutation-mediated resistance mechanisms in P. aeruginosa, like those leading to the inactivation of the carbapenem porin OprD, hyperproduction of the chromosomal cephalosporinase AmpC, or upregulation of RND efflux systems, are more frequently selected during treatment, leading to the failure of antimicrobial therapy (4, 10–12).

The combination of many of these mechanisms has led to the emergence of MDR or even extensively drug-resistant (XDR) strains, which are resistant to most of the currently available β-lactams. These strains are responsible for a growing number of outbreaks in the hospital setting associated with significantly higher morbidity and mortality attributed to limited effective antimicrobial options (7, 12, 13). Of further concern, recent studies have revealed that MDR/XDR P. aeruginosa clones (such as ST111, ST175, and ST235) have disseminated in multiple institutions worldwide, and for that reason, they have been classified as epidemic high-risk clones (14–16).

The multiplicity of resistance mechanisms, including MBLs, in P. aeruginosa poses a significant therapeutic challenge since even newer β-lactam–β-lactamase inhibitor (BL-BLI) combinations, particularly ceftolozane-tazobactam and ceftazidime-avibactam, are unable to provide therapeutic coverage for infections caused by such pathogens. The effectiveness of traditional approaches using BL-BLI combinations could continue to be challenged by the vast repertoire of mutations causing resistance in P. aeruginosa in the future as well (17–21).

Zidebactam (ZID) and WCK 5153 (Fig. 1) are the first described Gram-negative β-lactam enhancers belonging to the bicyclo-acyl hydrazide (BCH) series. ZID in combination with cefepime (FEP) is currently under clinical development for infections caused by MDR Gram-negative organisms, including P. aeruginosa and Acinetobacter baumannii. Although derived from a diazabicyclooctane (DBO) scaffold, BCHs were designed with the objective of augmenting PBP2 binding in P. aeruginosa and A. baumannii rather than the conventional approach of optimizing the β-lactamase-inhibitory activity of the compound. Avibactam, the first example of a DBO, possessed weak PBP2 affinity in Enterobacteriaceae, followed by OP0595 (RG 6080), which showed improved PBP2 affinity; however, the activity was limited to Enterobacteriaceae (22–24). In this work, we show that ZID and WCK 5153 represent an “advanced generation” of β-lactam enhancers with activity spectra targeted toward clinically important Gram-negative pathogens, including MDR/XDR P. aeruginosa.

FIG 1.

Chemical structures of ZID (zidebactam; WCK 5107) and WCK 5153.

The objective of this study was to characterize the mechanism of action of ZID and WCK 5153 and to investigate their activities alone and in combination with several different β-lactams against a well-characterized collection of P. aeruginosa strains expressing the most relevant β-lactam resistance determinants (e.g., AmpC hyperproduction, porin loss [oprD], or MexAB-OprM efflux pump overexpression) (14, 25–28). Furthermore, two well-characterized P. aeruginosa clinical isolates belonging to XDR MBL-producing epidemic high-risk clones, ST111 (VIM-1) and ST175 (VIM-2), were also tested (14).

RESULTS

MICs and minimum bactericidal concentrations (MBCs) of comparator β-lactams, zidebactam, WCK 5153, and their combinations with cefepime against P. aeruginosa PAO1 and knockout strains are shown in Table 1. The P. aeruginosa PAO1 MICs were 2 μg/ml for WCK 5153 and 4 μg/ml for zidebactam. Zidebactam and WCK 5153 MICs for the AmpC β-lactamase-hyperproducing derivatives remained within 1 doubling dilution, suggestive of low-level to no class C β-lactamase hydrolysis. Likewise, neither overexpression nor lack of the intrinsic efflux pump MexAB-OprM caused a MIC change of more than 1 doubling dilution. Furthermore, zidebactam and WCK 5153 MBCs remained within 1 doubling dilution of the MIC for nearly all tested strains, reflecting their inherent bactericidal activity. Cefepime MICs and MBCs decreased by 2 to 5 doubling dilutions when the drug was combined with subinhibitory concentrations of zidebactam or WCK 5153, demonstrating a noticeable inhibitory effect by zidebactam and WCK 5153 in nearly all tested P. aeruginosa strains.

TABLE 1.

MICs and MBCs of β-lactams and zidebactam and WCK 5153 in the studied strains

Straina	MIC/MBC (μg/ml)b,c,d
	FEP	MEM	MEC	ZID	WCK 5153	FEP + ZID (2 μg/ml)	FEP + WCK 5153 (1 μg/ml)	FEP + ZID (1 μg/ml)	FEP + WCK 5153 (0.5 μg/ml)
PAO1	1/2	0.25/0.5	>32/ND	4/8	2/4	0.03/0.12	0.03/0.06	0.06/0.12	0.06/0.06
PAdB	16/32	2/2	>32/ND	8/8	4/8	4/4	2/2	4/8	4/4
PAΔDDh2Dh3	16/16	1/2	>32/ND	4/16	2/8	1/2	1/4	2/8	2/4
PAOD	1/2	2/2	>32/ND	4/8	2/4	0.06/0.12	0.03/0.06	0.25/0.5	0.06/0.12
PAOM	0.12/0.12	0.06/0.12	2/ND	2/2	2/2	0.03/0.03	0.03/0.06	0.06/0.06	0.03/0.6
PAOMxR	2/4	1/2	>32/ND	8/16	4/4	2/4	2/2	4/4	2/2

^aPAO1, wild-type reference strain; PAdB, dacB knockout mutant of PAO1; PAΔDDh2Dh3, ampD triple (ampD-ampDh2-ampDh3) knockout mutant of PAO1; PAOD, PAO1 oprD-defective mutant of the porin OprD; PAOM, oprM knockout mutant of PAO1; PAOMxR, mexR knockout mutant of PAO1.

^bFEP, cefepime; MEM, meropenem; MEC, amdinocillin; ZID, zidebactam; WCK 5153, bicyclo-acyl hydrazide.

^cClinical and Laboratory Standards Institute (CLSI) susceptibility breakpoints: FEP, ≤8 μg/ml; MEM, ≤2 μg/ml; MEC, ZID, and WCK 5153, not determined (ND).

^dRange of concentrations tested, 0.0156 to 32 μg/ml.

The 50% inhibitory concentrations (IC₅₀s) (mean ± standard deviation from at least 3 independent experiments) of zidebactam, WCK 5153, cefepime, meropenem, and amdinocillin for the P. aeruginosa PAO1 PBPs are displayed in Table 2. Zidebactam and WCK 5153 were found to have high and exclusive PBP2 affinity, showing an inhibition similar to that obtained with amdinocillin. On the other hand, cefepime showed potent PBP1a and PBP3 inhibition, while meropenem inhibited PBP2, PBP3, and PBP4.

TABLE 2.

IC₅₀s of cefepime, meropenem, amdinocillin, zidebactam, and WCK 5153 for P. aeruginosa PAO1 PBPs

PBP	IC50 (μg/ml)a
	FEP	MEM	MEC	ZID	WCK 5153
1a	0.12 ± 0.01	0.26 ± 0.08	>4	>4	>4
1b	0.82 ± 0.07	0.21 ± 0.10	>4	>4	>4
2	2.7 ± 0.9	0.13 ± 0.02	0.19 ± 0.02	0.26 ± 0.06	0.14 ± 0.05
3	0.15 ± 0.07	0.06 ± 0.03	>4	>4	>4
4	2.5 ± 0.3	0.01 ± 0.003	>4	>4	>4
5/6	>4	>4	>4	>4	>4

^aMean values ± standard deviations from at least 3 independent experiments are shown.

Zidebactam and WCK 5153 were previously shown to inhibit class A and C β-lactamases, even enzymes with high hydrolytic activities, such as PDC-3 (extended-spectrum P. aeruginosa AmpC [ESAC] β-lactamase) or KPC-2 (29). Nevertheless, none of the compounds demonstrated significant class D enzyme inhibition (30). Our VIM-2 inhibition assay undertaken with purified enzyme revealed that zidebactam and WCK 5153 did not inhibit VIM-2, as evidenced by the high apparent K_i (K_{i app}) value (>100 μM).

Time-kill studies performed on wild-type strain PAO1 at a large inoculum (≥10⁷) are shown in Fig. 2. Zidebactam first displayed a bacteriostatic effect through all the concentrations tested (0.25×, 1×, and 2× MIC). Furthermore, when 4× MIC of WCK 5153 was applied, a close to 3-log-higher bactericidal activity was shown. In combination with 1× MIC of cefepime, zidebactam and WCK 5153 showed enhanced bactericidal effects at concentrations as low as 0.25× MIC (1 or 0.5 μg/ml), with an ≈3-log reduction of bacterial load during the first 8 h of incubation. Moreover, bactericidal activity was further increased when 1× MIC of either zidebactam or WCK 5153 was used.

FIG 2.

Killing curves measured in terms of reduction of viable CFU per milliliter over time for wild-type strain PAO1 at a 10⁷ inoculum. The concentrations tested were 1 and 4 μg/ml of cefepime (FEP) (1× and 4× MIC); 1, 4, and 8 μg/ml of zidebactam (ZID) (0.25×, 1×, and 2× MIC) and WCK 5153 (0.5×, 2×, and 4× MIC); and combinations of 1 and 4 μg/ml of FEP (1× and 4× MIC) with 0.25× to 1× MICs of ZID or WCK 5153. Mean values for three experiments ± the standard deviations are shown.

Consistent with PBP assays and time-kill curves, cefepime inhibition of PBP3 led to pronounced filamentation of the cells through different times and concentrations (1× and 4× MIC) (Fig. 3). Likewise, inhibition of PBP3 by the lower meropenem concentration (1× MIC) caused filamentation; however, when the higher concentration was tested (4× MIC), spindle-shaped cells were formed as a result of simultaneous PBP2 and PBP3 inhibition (31). The inhibitory activity of zidebactam and WCK 5153 toward PBPs resulted in the formation of spheroplasts attributable to the exclusive binding of PBP2 (Fig. 3). Consequently, when combining zidebactam or WCK 5153 with cefepime, spindle-shaped cells like those observed with meropenem were observed to result from the concomitant inhibition of PBP2 and PBP3. LIVE-DEAD staining further supported the mechanistic basis of bactericidal effect of both zidebactam and WCK 5153 enhancer combinations on the P. aeruginosa PAO1 strain.

FIG 3.

Results of LIVE/DEAD staining performed on wild-type PAO1. (A) Images obtained at 2 h of incubation with 1× MICs of cefepime (FEP), meropenem (MEM), zidebactam (ZID), and WCK 5153 and with 1× to 0.25× MIC of cefepime-zidebactam or cefepime-WCK 5153. (B) Images obtained at 8 h of incubation with 4× MICs of cefepime, zidebactam, and WCK 5153 and with 4× to 0.5× MICs of cefepime-zidebactam or cefepime-WCK 5153. Live cells are stained green, and dead cells are stained red.

Additional susceptibility testing with an expanded panel of β-lactams and combinations using cefepime, meropenem, and amdinocillin was performed. Aztreonam, piperacillin, imipenem, doripenem, zidebactam, and WCK 5153 testing against two MBL-producing clinical strains was conducted. Despite the MICs of ZID and WCK 5153 being higher for the MBL-producing clinical strains (16 to 32 μg/ml) than those for PAO1 and derivative mutants, enhancement in the partner β-lactam activity was noted at both concentrations (4 and 8 μg/ml) for both MBL-producing isolates.

Figure 4 shows the results of time-kill experiments for cefepime, aztreonam, and piperacillin using CLSI breakpoint concentrations (susceptible and intermediate) alone or in combination with 8 μg/ml of zidebactam or WCK 5153 against an AmpC-hyperproducing P. aeruginosa PAO1 dacB mutant. A significant improvement in bactericidal activity (>3-log reduction compared to individual regimens) was observed, leading to complete eradication of bacterial cultures at 24 h, showing that these compounds are not hydrolyzed by class C AmpC β-lactamase.

FIG 4.

Results of the killing curves measured in terms of reduction of viable CFU per milliliter over time for the AmpC-hyperproducing PAO1 dacB mutant (PAdB). The concentrations tested were 8 and 16 μg/ml of cefepime (FEP) (0.5× and 1× MIC), 8 and 16 μg/ml of aztreonam (ATM) (0.03× and 0.06× MIC), and 16 and 32 μg/ml of piperacillin (PIP) (0.06× and 0.125× MIC), at susceptible and intermediate CLSI breakpoint concentrations in combination with 8 μg/ml of zidebactam (ZID) or WCK 5153. Standalone drugs were tested at the maximum concentration used for combinations. Mean values for three experiments ± the standard deviations are shown. The dashed line represents the limit of detection.

MexAB-OprM-hyperproducing PAO1 mexR mutant killing curves are shown in Fig. 5. Zidebactam and WCK 5153, as previously seen in MIC determinations, possessed a lower bactericidal activity than with the AmpC-hyperproducing strain. Furthermore, combinations with cefepime or aztreonam displayed a more concentration-dependent enhancer effect, suggesting that both compounds could be slightly affected by the MexAB-OprM efflux system. An OprM knockout mutant displayed MICs and MBCs that were 1 to 2 dilutions lower, thus further supporting that MexAB-OprM might play a role in zidebactam and WCK 5153 tolerance.

FIG 5.

Results of the killing curves measured in terms of reduction of viable CFU per milliliter over time for the MexAB-OprM-hyperproducing PAO1 mexR mutant (PAOMxR). Cefepime (FEP) (8 and 16 μg/ml), aztreonam (ATM) (8 and 16 μg/ml), piperacillin (PIP) (16 and 32 μg/ml), imipenem (IPM) (4 and 8 μg/ml), meropenem (MEM) (4 and 8 μg/ml), and doripenem (DOR) (4 and 8 μg/ml) were tested at susceptible and intermediate CLSI breakpoint concentrations in combination with 8 μg/ml of zidebactam (ZID) or WCK 5153. Standalone drugs were tested at the maximum concentration used for combinations. Mean values for three experiments ± the standard deviations are shown. The dashed line represents the limit of detection.

To further evaluate these combinations against clinical isolates of MDR P. aeruginosa, time-kill studies were undertaken with XDR VIM-1 producer P. aeruginosa clone ST111 (Fig. 6) and VIM-2 producer P. aeruginosa clone ST175 (Fig. 7). We found that all of the antipseudomonal β-lactams displaying high PBP3 affinity (cefepime, piperacillin, imipenem, meropenem, and doripenem) and the sole PBP3 binder aztreonam (susceptible and intermediate CLSI breakpoint concentrations) tested in combination with 8 μg/ml of ZID or WCK 5153 led to potent synergistic killing (≥2 to 3 logs) for the VIM-1- and VIM-2-producing isolates. MICs for these strains are shown in Table 3.

FIG 6.

Results of the killing curves measured in terms of reduction of viable CFU per milliliter over time for the XDR ST111 (VIM-1) isolate. Cefepime (FEP) (8 and 16 μg/ml), aztreonam (ATM) (8 and 16 μg/ml), piperacillin (PIP) (16 and 32 μg/ml), imipenem (IPM) (4 and 8 μg/ml), meropenem (MEM) (4 and 8 μg/ml), and doripenem (DOR) (4 and 8 μg/ml) were tested at susceptible and intermediate CLSI breakpoint concentrations in combination with 8 μg/ml of zidebactam (ZID) or WCK 5153. Standalone drugs were tested at the maximum concentration used for combinations. Mean values for three experiments ± the standard deviations are shown. The dashed line represents the limit of detection.

FIG 7.

Results of the killing curves measured in terms of reduction of viable CFU per milliliter over time for the XDR ST175 (VIM-2) isolate. Cefepime (FEP) (8 and 16 μg/ml), aztreonam (ATM) (8 and 16 μg/ml), piperacillin (PIP) (16 and 32 μg/ml), imipenem (IPM) (4 and 8 μg/ml), meropenem (MEM) (4 and 8 μg/ml), and doripenem (DOR) (4 and 8 μg/ml) were tested at susceptible and intermediate CLSI breakpoint concentrations in combination with 8 μg/ml of zidebactam (ZID) or WCK 5153. Standalone drugs were tested at the maximum concentration used for combinations. Mean values for three experiments ± the standard deviations are shown. The dashed line represents the limit of detection.

TABLE 3.

MICs of β-lactams, zidebactam, WCK 5153, and combinations for the high-risk MBL-producing strains tested in killing kinetics studies

Antibiotic/NCEa,b	MIC (μg/ml)c,d
	ST111 (VIM-1)	ST175 (VIM-2)
FEP	512	32
FEP + ZID (4 μg/ml)	16	4
FEP + ZID (8 μg/ml)	2	≤0.5
FEP + WCK 5153 (4 μg/ml)	4	1
FEP + WCK 5153 (8 μg/ml)	≤0.5	≤0.5
ATM	32	128
ATM + ZID (4 μg/ml)	2	≤0.5
ATM + ZID (8 μg/ml)	≤0.5	≤0.5
ATM + WCK 5153 (4 μg/ml)	≤0.5	≤0.5
ATM + WCK 5153 (8 μg/ml)	≤0.5	≤0.5
PIP	2,048	64
PIP + WCK 5153 (4 μg/ml)	4	32
PIP + WCK 5153 (8 μg/ml)	≤0.5	16
IPM	256	64
IPM + WCK 5153 (4 μg/ml)	8	16
IPM + WCK 5153 (8 μg/ml)	≤0.5	8
MEM	256	16
MEM + WCK 5153 (4 μg/ml)	2	≤0.5
MEM + WCK 5153 (8 μg/ml)	≤0.5	≤0.5
DOR	128	16
DOR + WCK 5153 (4 μg/ml)	4	≤0.5
DOR + WCK 5153 (8 μg/ml)	≤0.5	≤0.5
ZID	32	32
WCK 5153	16	16

^aFEP, cefepime; ATM, aztreonam; PIP, piperacillin; IPM, imipenem; MEM, meropenem; DOR, doripenem; ZID, zidebactam; WCK 5153, bicyclo-acyl hydrazide; NCE, new chemical entity.

^bClinical and Laboratory Standards Institute (CLSI) susceptibility breakpoints: FEP and ATM, ≤8 μg/ml; PIP, ≤16 μg/ml; IPM, MEM, and DOR, ≤2 μg/ml; and ZID and WCK 5153, not applicable.

^cST111 (VIM-1), XDR VIM-1 producer P. aeruginosa high-risk ST111 clone (E-XDR-ST111-1-VIM1); ST175 (VIM-2), XDR VIM-2 producer P. aeruginosa high-risk ST175 clone (E-XDR-ST175-17-VIM2).

^dRange of concentrations tested, 0.5 to 2,048 μg/ml.

DISCUSSION

P. aeruginosa is characterized by a complex repertoire of natural antibiotic resistance elements resulting in intrinsic resistance to many antibiotics and lower overall susceptibility to nearly all agents. Moreover, the remarkable ability for developing high-level resistance (to nearly all available antibiotics) through the selection of chromosomal mutations makes the treatment of P. aeruginosa infections quite complex. As a result, the treatment of infections by MDR/XDR strains is even more daunting (4, 5, 32).

For the treatment of MDR/XDR P. aeruginosa, combination therapy using β-lactams together with an aminoglycoside, a fluoroquinolone, or colistin is frequently used in an attempt to anticipate and overcome resistance. Regrettably, combination therapy is sometimes fraught with toxicity (14, 33–39). The recent introduction of new BL-BLI combinations, particularly ceftolozane-tazobactam and ceftazidime-avibactam, into clinical practice is helping to alleviate to some extent the current medical need. However, these combinations may not fully escape from the vast repertoire of P. aeruginosa resistance mutations and are not active against MBL-producing strains that are increasingly reported (17–21).

In this work, we provide the first evidence that the new DBO derivatives (zidebactam and WCK 5153) display antipseudomonal activity driven by potent PBP2 inhibition and, being of a new antibacterial chemotype, are not significantly affected by classical P. aeruginosa resistance mechanisms. This significantly expands recent findings for OP0595 showing PBP2-related activity against Enterobacteriaceae only (23). Moreover, a marked additive effect with most β-lactams, particularly those strongly inhibiting PBP1 and PBP3, such as cefepime, was observed for wild-type P. aeruginosa PAO1, thus laying the foundation for the “β-lactam enhancer” concept.

It has been reported that inhibition of a single PBP leads to either a bacteriostatic effect or a low rate of killing, whereas saturation of two of the three essential PBPs leads to improvement of bactericidal action, which reaches to the maximum extent of killing when PBP1a, -1b, -2, and -3 are concomitantly inhibited (27, 40, 41). This is likely the principle of the potent β-lactam enhancer-mediated killing observed for zidebactam (PBP2 IC₅₀: 0.26 μg/ml) and WCK 5153 (PBP2 IC₅₀: 0.14 μg/ml) combined with cefepime, a potent PBP1a and PBP3 binder.

Subinhibitory concentrations of zidebactam and WCK 5153 enhanced the susceptibility of the wild-type strain to cefepime. Likewise, this enhancer effect was observed for strains harboring clinically relevant β-lactam-impacting resistance mutations, such as those leading to high levels of AmpC hyperproduction or porin loss (oprD). Moreover, time-kill experiments performed with AmpC and efflux pump-hyperproducing strains revealed sustained bactericidal activity (>3-log reduction compared to those with individual regimens) of combinations, up to 24 h. Thus, these results indicate that zidebactam and WCK 5153 could be valuable β-lactam partners to combat mutational resistance in P. aeruginosa.

It is worth mentioning that against XDR MBL-producing epidemic high-risk clones, zidebactam and WCK 5153 in combination with different β-lactams, at clinically attainable concentrations, led to potent killing (≥3 logs). Aztreonam showed a rather more concentration-dependent killing than did cefepime, displaying a lower additive effect. This observation is attributable to the fact that aztreonam binds solely PBP3.

Combinations of doripenem and meropenem with either zidebactam or WCK 5153 showed higher enhancer effects than the combinations of these compounds with imipenem. This is another hint toward the importance of PBP occupancy when rationally designing β-lactam combinations, since as previously described, this effect represents a direct consequence of the higher binding affinity for PBP2 and -3 of doripenem and meropenem in P. aeruginosa (42).

The basis of high bactericidal action emanates from unhindered high-affinity PBP2 engagement by zidebactam and WCK 5153, which causes the cells to rapidly convert into spheroplasts regardless of the linked resistance mechanisms. This could be the outcome of optimal permeation features of zidebactam and WCK 5153 coupled with their intrinsic MBL stability. Presumably, spheroplast formation causes perturbation in the outer membrane, leading to modulation of membrane-bound resistance mechanisms such as efflux, porin, and expression of β-lactamases (43–45). A similar phenomenon has been observed for amdinocillin-induced spheroplasts where leakage of β-lactamases has been reported (46). Under such scenarios, the concurrent presence of cefepime results in complementary engagement of other essential PBPs, leading to the onset of pronounced bacterial lysis.

In summary, zidebactam and WCK 5153 in combination with β-lactams are the first Gram-negative β-lactam enhancer-based agents providing coverage of MDR/XDR P. aeruginosa, even for those strains expressing metallo-β-lactamases.

MATERIALS AND METHODS

Laboratory and clinical strains.

PAO1 knockout mutations in genes resulting in the increased expression of the chromosomal β-lactamase AmpC (i.e., a triple ampD knockout [PAΔDDh2Dh3] and a dacB knockout [PAΔdB]), in the increased or decreased expression, respectively, of the MexAB-OprM efflux pump (i.e., a mexR knockout [PAOMR] and an oprM knockout [PAΔOM]), and in the loss of the outer membrane porin OprD (PAΔOD) were obtained from previous studies (26, 28, 47). Additionally, two well-characterized MBL-producing XDR clinical isolates belonging to the epidemic high-risk clones ST111 (VIM-1) and ST175 (VIM-2) were used (Table 4) (14).

TABLE 4.

Strains used in this study

Strain	Genotype/relevant characteristics	Reference
PAO1	Reference strain completely sequenced	52
PAdB	PAO1 ΔdacB::lox.dacB; encodes the nonessential PBP4	26
PAΔDDh2Dh3	PAO1 ΔampD::lox ΔampDh2::lox ΔampDh3::lox; AmpD, AmpDh2 and AmpDh3 are the three N-acetyl-amidases of P. aeruginosa	25
PAOD	PAO1 oprD-G194A (W65X) defective mutant of the porin OprD	47
PAOM	PAO1 ΔoprM::lox oprM; encodes the outer membrane protein component of MexAB-OprM and MexXY-OprM efflux pumps	28
PAOMxR	PAO1 ΔmexR::lox mexR; encodes the negative regulator of MexAB-OprM efflux pump	28
ST111 (VIM-1)	XDR VIM-1 producer P. aeruginosa high-risk ST111 clone (E-XDR-ST111-1-VIM1)	14
ST175 (VIM-2)	XDR VIM-2 producer P. aeruginosa high-risk ST-175 clone (E-XDR-ST175-17-VIM2)	14

Susceptibility testing.

The MICs and the minimum bactericidal concentrations (MBCs) of cefepime, meropenem, amdinocillin, zidebactam, and WCK 5153 as well as the combinations of cefepime-zidebactam and cefepime-WCK 5153 were determined for wild-type strain PAO1 by following the standard Clinical and Laboratory Standards Institute (CLSI) broth microdilution method (48).

The MICs of cefepime, meropenem, amdinocillin, aztreonam, piperacillin, imipenem, doripenem, and the combinations cefepime-zidebactam, cefepime-WCK 5153, aztreonam-zidebactam, aztreonam-WCK 5153, piperacillin-WCK 5153, imipenem-WCK 5153, meropenem-5153, and doripenem-WCK 5153 were determined for the MBL-producing XDR clinical isolates belonging to the epidemic high-risk clones ST111 (VIM-1) and ST175 (VIM-2) by following the standard Clinical and Laboratory Standards Institute (CLSI) broth microdilution method (48). Zidebactam, WCK 5153, and comparator/partner β-lactams were provided by Wockhardt Ltd.

Time-kill kinetics.

For killing kinetics studies, overnight Mueller-Hinton broth (MHB) cultures of PAO1, PAΔdB, PAO1ΔmexR, ST111 (VIM-1), and ST175 (VIM-2) were diluted (1/100) in fresh medium and incubated at 37°C with shaking (180 rpm). Cultures were grown to an optical density at 600 nm (OD₆₀₀) of 0.2 (early-log-phase growth). Killing curves for the wild-type, susceptible PAO1 strain were then initiated by inoculating microtiter plates containing MHB (initial inoculum, 1 × 10⁷ to 5 × 10⁷ CFU/ml) in the presence of 1 and 4 μg/ml of cefepime (1× and 4× MIC), 1, 4, and 8 μg/ml of zidebactam (0.25×, 1×, and 2× MIC), and WCK 5153 (0.5×, 2×, and 4× MIC) and combinations of 1 and 4 μg/ml of cefepime (1× and 4× MIC) with 0.25× to 1× MICs of zidebactam or WCK 5153.

Killing kinetics studies for the antibiotic-resistant PAO1 knockout mutants and MDR clinical isolates were initiated by inoculating 1 × 10⁶ to 5 × 10⁶ CFU/ml containing MHB in the presence of clinically relevant concentrations of cefepime (8 and 16 μg/ml), aztreonam (8 and 16 μg/ml), and piperacillin (16 and 32 μg/ml) (susceptible and intermediate CLSI breakpoint concentrations, respectively) in combination with zidebactam or WCK 5153 at 8 μg/ml.

Killing kinetics were monitored during a 24-h period and were analyzed by CFU enumeration (plating serial dilutions on Mueller-Hinton agar). In addition, LIVE-DEAD staining was conducted with a LIVE/DEAD BacLight bacterial viability kit (Molecular Probes, Invitrogen, Carlsbad, CA) for the wild-type PAO1 strain by following the manufacturer's instructions and using a Nikon Eclipse E400 fluorescence microscope at a magnification of ×1,000. All experiments were performed in triplicate.

PBP assays.

Membrane preparations containing the PBPs of P. aeruginosa PAO1 were obtained by following previously described protocols (27, 49). Briefly, 400 ml of late-log-phase (OD₆₀₀ = 1) cultures of PAO1 were collected by centrifugation, washed, and resuspended in 20 mM KH₂PO₄ with140 mM NaCl pH 7.5 (buffer A). Cells were then sonicated and centrifuged at 4,000 × g for 20 min.

The bacterial membranes were collected by ultracentrifugation, and total protein content was measured using the Bradford method. Twenty micrograms of P. aeruginosa PBP-containing membrane preparation was then incubated for 30 min at 35°C in the presence of increasing concentrations of cefepime, meropenem, zidebactam, and WCK 5153 (range of concentrations tested: 0.0156 to 2 μg/ml) and were afterwards labeled with a 25 μM concentration of the fluorescent penicillin Bocillin FL (49). Labeled PBPs were separated through 10% SDS-polyacrylamide gel electrophoresis (Bio-Rad Laboratories, Hercules, CA). Labeled PBPs were visualized using a Typhoon FLA 9500 biomolecular imager (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) (excitation at 488 nm and emission at 530 nm), and 50% inhibitory concentrations (IC₅₀s) for the different PBPs were determined from at least three independent membrane preparations using ImageQuant TL software (GE Healthcare Bio-Sciences AB).

VIM-2 purification and inhibition by zidebactam.

The VIM-2 β-lactamase was purified from Escherichia coli BL21(DE3) pLys cells carrying pET-24a(+)-bla_VIM-2-S21, as previously described (50). The inhibition of VIM-2 by ZID was analyzed by determining the apparent K_i (K_{i app}) value using an Agilent (Santa Clara, CA) 8453 diode array spectrophotometer. All reactions were conducted in 10 mM phosphate-buffered saline (PBS) at pH 7.4 and room temperature. A direct competition assay between nitrocefin (reporter substrate) and zidebactam with VIM-2 was conducted using pseudo-first-order conditions applying steady-state kinetics. The three reaction components were mixed manually, and the reaction velocity for the first 10 s of the reaction was collected. The data were linearized by plotting the inverse initial steady-state velocities (1/V₀) versus zidebactam concentration. K_{i app} was determined by dividing the value for the y-intercept by the slope of the line and correcting for the use of nitrocefin, as previously described (51).